agentsclimarketplace

Case 02592

Skill knownasnaffy/prompthound/dataset/case_02592

A fast, offline static risk analysis CLI for AI agent skill files. Detects malicious instructions, steganographic payloads, and dangerous capability chains.

Install
npx -y skills add knownasnaffy/prompthound --skill case_02592

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Lead generation, prospecting, and qualification for B2B sales. Use when asked to find leads, build prospect lists, research target companies, qualify leads against an ICP, enrich contact data, find decision-maker email formats, scrape company info, score leads, or build a sales pipeline from scratch. Triggers on phrases like "find leads", "build a prospect list", "who should I target", "lead generation", "find decision makers", "ICP research", "sales prospecting", "qualify leads", "enrich contacts", "find emails", "pipeline building".

SKILL.md

2.9 KB, 574 tokens by cl100k_base, as published. Nobody here has run it

Lead Generation Skill

Build targeted, qualified prospect lists and enrich them with decision-maker data. Output is pipeline-ready — not raw noise.

Workflow

1. Define the ICP (Ideal Customer Profile)

Ask if not provided:

  • Industry/vertical: (e.g. SaaS, e-commerce, professional services)
  • Company size: (employees or revenue range)
  • Geography: (country, region, or global)
  • Job titles to target: (e.g. Head of Marketing, VP Sales, Founder)
  • Pain point or trigger event: (e.g. recently funded, hiring for X role, using competitor Y)

2. Build the Target Company List

Use web_search with precision queries. See references/search-playbook.md for query templates.

Effective discovery sources:

  • LinkedIn Sales Navigator signals (via search queries)
  • G2/Capterra category pages (companies using specific software)
  • Crunchbase (funded companies by stage, industry, date)
  • Job boards (companies hiring = growing = buying)
  • Industry directories and association member lists
  • Subreddits and communities where ICP hangs out

3. Find Decision Makers

For each target company:

  • Search "[company] [job title] LinkedIn" to identify names
  • Use web_search to find personal/professional profiles
  • Cross-reference company About/Team pages

4. Enrich Contacts

For each contact, gather:

  • Full name, title, company
  • LinkedIn URL
  • Email format (guess from pattern: [email protected], [email protected])
  • Company size, industry, location
  • Recent trigger events (new role, funding, product launch)

Run scripts/enrich_leads.py to format and score the list.

5. Score & Prioritize

Score each lead 1–10 using the rubric in references/scoring-rubric.md:

  • ICP fit (industry, size, title match)
  • Buying signals (trigger events, tech stack, intent)
  • Reachability (email confidence, LinkedIn activity)
  • Timing (recently funded, new hire, Q1 budget cycle)

6. Output Format

Deliver as a CSV-ready table + summary:

NameTitleCompanyIndustrySizeEmail (est.)LinkedInScoreNotes

Include:

  • Total leads found
  • Score distribution
  • Top 10 "strike now" leads highlighted
  • Suggested outreach angle per segment

See references/search-playbook.md for advanced search query patterns. See references/email-formats.md for common company email format patterns.

What ships with it: 3 files

11.8 KB alongside SKILL.md, 1 of them executable

references/

scripts/

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.